Sprague-Dawley rat hearts were perfused under constant flow conditions, and a balloon was inserted into the left ventricle to measure heart rate (HR) and left ventricular pressures. Left ventricular generated pressure (LVGP) was calculated as peak systolic minus end diastolic pressure. Three substrate groups, pyruvate (5 mM), glucose (15 mM) and octanoate (0.5 mM), were employed. Oxidative stress was induced by perfusion with tertiary-butyl hydroperoxide (tBHP, 0.35 mM, 12 min) followed by 25 min of perfusion with control buffer. Hearts perfused with pyruvate showed no significant decrease in contractile function following tBHP treatment (HR x LVGP = 17666 +/- 585 mmHg/min, initial: 16414 +/- 2083 post-tBHP treatment). Glucose-perfused hearts had an intermediate decrease in function (19174 +/- 828 mmHg/min, initial; 4379 +/- 2083 post-tBHP), while octanoate-perfused hearts recovered no contractile function. Peak release of LDH was lowest in hearts perfused with pyruvate (115 +/- 17 mU/g wet wt/min), intermediate in glucose-perfused hearts (1575 +/- 380) and highest in octanoate-perfused hearts (3074 +/- 499). Thiobarbituric acid reactive substances (TBARS) were unchanged in hearts perfused with pyruvate (16.2 +/- 5 nmoles/g wet wt), but increased significantly in glucose-perfused hearts (36.1 +/- 1) and in octanoate-perfused hearts (45.5 +/- 9). Total glutathione levels were unchanged in hearts perfused with pyruvate (753 +/- 68 nmoles/g wet wt), but significantly decreased in glucose-perfused hearts (594 +/- 68) and in octanoate-perfused hearts (445 +/- 38) following tBHP-treatment. Pyruvate significantly reduced oxidative injury. In contrast, glucose provided a small reduction in injury while octanoate-perfused hearts had the most severe injury.(ABSTRACT TRUNCATED AT 250 WORDS)